def test_process_with_aria(self):
"""Demonstrates advanced usage considering accessibility."""
expected_chunks = [
budou.Chunk(u'今日は', u'NOUN', u'NN', True),
budou.Chunk(u'晴れ。', u'NOUN', u'ROOT', False)
]
expected_html_code = (
u'<span aria-describedby="parent" class="text-chunk">今日は</span>'
u'<span aria-describedby="parent" class="text-chunk">晴れ。</span>')
result = self.parser.parse(DEFAULT_SENTENCE, {
'aria-describedby': 'parent',
'class': 'text-chunk'
}, use_cache=False)
self.assertIn(
'chunks', result,
'Processed result should include chunks.')
self.assertIn(
'html_code', result,
'Processed result should include organized html code.')
self.assertEqual(
expected_chunks, result['chunks'],
'Processed result should include expected chunks.')
self.assertEqual(
expected_html_code, result['html_code'],
'Processed result should include expected html code.')
def setUp(self):
queue = budou.ChunkQueue()
chunks = [
budou.Chunk('ab', dependency=None),
budou.Chunk('cde', dependency=True),
budou.Chunk('fgh', dependency=False)
]
for chunk in chunks:
queue.add(chunk)
self.queue = queue
def test_get_source_chunks(self):
expected = [
budou.Chunk(u'今日', u'NOUN', u'NN', True),
budou.Chunk(u'は', u'PRT', u'PRT', False),
budou.Chunk(u'晴れ', u'NOUN', u'ROOT', False),
budou.Chunk(u'。', u'PUNCT', u'P', False),
]
result = self.parser._get_source_chunks(DEFAULT_SENTENCE)
self.assertEqual(
expected, result,
'Input sentence should be processed into source chunks.')
def test_get_chunks_per_space(self):
source = 'a b'
expected = [
budou.Chunk('a', None, None, True),
budou.Chunk(' ', budou.SPACE_POS, budou.SPACE_POS, True),
budou.Chunk('b', None, None, True)
]
result = self.parser._get_chunks_per_space(source)
self.assertEqual(
result, expected,
'Input text should be parsed into chunks separated by spaces.')
def test_spanize(self):
chunks = [
budou.Chunk(u'a', None, None, None),
budou.Chunk(u'b', None, None, None),
budou.Chunk(u'c', None, None, None),
]
classname = 'foo'
expected = (u'<span class="foo">a</span>'
'<span class="foo">b</span>'
'<span class="foo">c</span>')
result = self.parser._spanize(chunks, classname)
self.assertEqual(result, expected,
'The chunks should be compiled to a HTML code.')
def test_concatenate_punctuations(self):
chunks = [
budou.Chunk(u'a', None, None, None),
budou.Chunk(u'b', u'PUNCT', None, None),
budou.Chunk(u'c', None, None, None),
]
expected_forward_concat = [
budou.Chunk(u'ab', None, None, None),
budou.Chunk(u'c', None, None, None),
]
result = self.parser._concatenate_punctuations(chunks)
self.assertEqual(result, expected_forward_concat,
'Punctuation marks should be concatenated backward.')
def test_html_serialize(self):
chunks = budou.ChunkList([
budou.Chunk('Hello'),
budou.Chunk.space(),
budou.Chunk(u'今天'),
budou.Chunk(u'天气'),
budou.Chunk(u'很好')
])
attributes = {'class': 'foo'}
expected = ('<span>'
'Hello '
u'<span class="foo">今天</span>'
u'<span class="foo">天气</span>'
u'<span class="foo">很好</span>'
'</span>')
result = self.parser._html_serialize(chunks, attributes, None)
self.assertEqual(result, expected,
'The chunks should be compiled to a HTML code.')
chunks = budou.ChunkList([
budou.Chunk('Hey<'),
budou.Chunk('<script>alert(1)</script>'),
budou.Chunk('>guys')
])
attributes = {'class': 'foo'}
expected = ('<span>'
'Hey<<script>alert(1)</script>>guys'
'</span>')
result = self.parser._html_serialize(chunks, attributes, None)
self.assertEqual(result, expected,
'HTML tags included in a chunk should be encoded.')
chunks = budou.ChunkList([
budou.Chunk(u'去年'),
budou.Chunk(u'インフルエンザに'),
budou.Chunk(u'かかった。')
])
attributes = {'class': 'foo'}
expected = ('<span>'
u'<span class="foo">去年</span>'
u'インフルエンザに'
u'<span class="foo">かかった。</span>'
'</span>')
result = self.parser._html_serialize(chunks, attributes, 6)
self.assertEqual(
result, expected,
'Chunks that exceed the max length should not be enclosed by a span.'
)
def test_swap(self):
old_chunks = self.queue.chunks[0:2]
new_chunk = budou.Chunk('ijk')
self.queue.swap(old_chunks, new_chunk)
expected = ['ijk', 'fgh']
self.assertEqual(expected, [chunk.word for chunk in self.queue.chunks],
'Old chunks should be replaced with the new chunk.')
def test_migrate_html(self):
source = u'こ<a>ちらを</a>クリック'
dom = html.fragment_fromstring(source, create_parent='body')
chunks = [
budou.Chunk(u'こちら', u'PRON', u'NSUBJ', True),
budou.Chunk(u'を', u'PRT', u'PRT', False),
budou.Chunk(u'クリック', u'NOUN', u'ROOT', False),
]
expected = [
budou.Chunk(u'こ<a>ちらを</a>', budou.HTML_POS, budou.HTML_POS, True),
budou.Chunk(u'クリック', u'NOUN', u'ROOT', False),
]
result = self.parser._migrate_html(chunks, dom)
self.assertEqual(
expected, result,
'The HTML source code should be migrated into the chunk list.')
def test_html_serialize(self):
chunks = budou.ChunkList([
budou.Chunk('a'),
budou.Chunk('b'),
budou.Chunk.space(),
budou.Chunk('c')
])
attributes = {'class': 'foo'}
expected = ('<span>'
'<span class="foo">a</span>'
'<span class="foo">b</span> '
'<span class="foo">c</span>'
'</span>')
result = self.parser._html_serialize(chunks, attributes)
self.assertEqual(result, expected,
'The chunks should be compiled to a HTML code.')
def test_spanize(self):
queue = budou.ChunkQueue()
chunks = [
budou.Chunk('a'),
budou.Chunk('b'),
budou.Chunk.space(),
budou.Chunk('c'),
]
for chunk in chunks:
queue.add(chunk)
attributes = {'class': 'foo'}
expected = ('<span class="foo">a</span>'
'<span class="foo">b</span> '
'<span class="foo">c</span>')
result = self.parser._spanize(queue, attributes)
self.assertEqual(result, expected,
'The chunks should be compiled to a HTML code.')
def test_html_serialize(self):
chunks = budou.ChunkList([
budou.Chunk('Hello'),
budou.Chunk.space(),
budou.Chunk(u'今天'),
budou.Chunk(u'天气'),
budou.Chunk(u'很好')
])
attributes = {'class': 'foo'}
expected = ('<span>'
'Hello '
u'<span class="foo">今天</span>'
u'<span class="foo">天气</span>'
u'<span class="foo">很好</span>'
'</span>')
result = self.parser._html_serialize(chunks, attributes)
self.assertEqual(result, expected,
'The chunks should be compiled to a HTML code.')
chunks = budou.ChunkList([
budou.Chunk('Hey<'),
budou.Chunk('<script>alert(1)</script>'),
budou.Chunk('>guys')
])
attributes = {'class': 'foo'}
expected = ('<span>'
'Hey<<script>alert(1)</script>>guys'
'</span>')
result = self.parser._html_serialize(chunks, attributes)
self.assertEqual(result, expected,
'HTML tags included in a chunk should be encoded.')
def test_concatenate_inner(self):
chunks = budou.ChunkList()
chunks.append(budou.Chunk('ab', dependency=None))
chunks.append(budou.Chunk('cde', dependency=True))
chunks.append(budou.Chunk('fgh', dependency=False))
chunks = self.parser._concatenate_inner(chunks, True)
self.assertEqual(['ab', 'cdefgh'], [
chunk.word for chunk in chunks
], 'Chunks should be concatenated if they depends on the following word.'
)
self.assertEqual(
[None, False], [chunk.dependency for chunk in chunks],
'Dependency should persist even if it\'s concatenated by others.')
chunks = self.parser._concatenate_inner(chunks, False)
self.assertEqual(['abcdefgh'], [
chunk.word for chunk in chunks
], 'Chunks should be concatenated if they depends on the previous word.'
)
def test_process(self):
expected_chunks = [
budou.Chunk(u'今日は', u'NOUN', u'NN', True),
budou.Chunk(u'晴れ。', u'NOUN', u'ROOT', False)
]
expected_html_code = (u'<span class="ww">今日は</span>'
u'<span class="ww">晴れ。</span>')
result = self.parser.parse(DEFAULT_SENTENCE)
self.assertIn('chunks', result,
'Processed result should include chunks.')
self.assertIn('html_code', result,
'Processed result should include organized html code.')
self.assertEqual(expected_chunks, result['chunks'],
'Processed result should include expected chunks.')
self.assertEqual(
expected_html_code, result['html_code'],
'Processed result should include expected html code.')
def test_parse_ja(self):
"""Demonstrates standard usage in Japanese."""
expected_chunks = [
budou.Chunk(u'今日は', u'NOUN', u'NN', True),
budou.Chunk(u'晴れ。', u'NOUN', u'ROOT', False)
]
expected_html_code = (u'<span class="ww">今日は</span>'
u'<span class="ww">晴れ。</span>')
result = self.parser.parse(DEFAULT_SENTENCE_JA,
language='ja',
use_cache=False)
self.assertEqual(
expected_chunks, result['chunks'],
'Processed result should include expected chunks in Japanese.')
self.assertEqual(
expected_html_code, result['html_code'],
'Processed result should include expected html code in Japanese.')
def test_parse_ko(self):
"""Demonstrates standard usage in Japanese."""
expected_chunks = [
budou.Chunk(u'오늘은', None, None, True),
budou.Chunk(' ', budou.SPACE_POS, budou.SPACE_POS, True),
budou.Chunk(u'맑음.', None, None, True)
]
expected_html_code = (u'<span class="ww">오늘은</span> '
u'<span class="ww">맑음.</span>')
result = self.parser.parse(DEFAULT_SENTENCE_KO,
language='ko',
use_cache=False)
self.assertEqual(
expected_chunks, result['chunks'],
'Processed result should include expected chunks in Korean.')
self.assertEqual(
expected_html_code, result['html_code'],
'Processed result should include expected html code in Korean.')
def test_maybe_add_dependency(self):
chunk = budou.Chunk('foo', label=None)
chunk.maybe_add_dependency(True)
self.assertEqual(
None, chunk.dependency,
'Dependency should not be added if the chunk does not belong to'
'dependent labels.')
chunk = budou.Chunk('foo', label=budou.Chunk.DEPENDENT_LABEL[0])
chunk.maybe_add_dependency(True)
self.assertEqual(
True, chunk.dependency,
'Dependency should be added if the chunk belongs to dependent labels.'
)
chunk = budou.Chunk('foo', label=budou.Chunk.DEPENDENT_LABEL[0])
chunk.dependency = False
chunk.maybe_add_dependency(True)
self.assertEqual(
False, chunk.dependency,
'Dependency should not be added if the chunk has dependency already.'
)
def test_add_dependency_if_punct(self):
test_characters = [
u'。', u'、', u'「', u'」', u'（', u'）', u'[', u']', u'(', u')'
]
expected_dependency = [
False, False, True, False, True, False, True, False, True, False
]
for i, character in enumerate(test_characters):
# _add_dependency_if_punct is called in __init__ implicitly.
chunk = budou.Chunk(character, pos='PUNCT')
self.assertEqual(
expected_dependency[i], chunk.dependency,
'Punctuation marks should be assigned with proper dependencies.'
)
def test_concatenate_by_label(self):
chunks = [
budou.Chunk(u'a', None, budou.TARGET_LABEL[0], True),
budou.Chunk(u'b', None, budou.TARGET_LABEL[1], False),
budou.Chunk(u'c', None, budou.TARGET_LABEL[2], True),
]
expected_forward_concat = [
budou.Chunk(u'ab', None, budou.TARGET_LABEL[1], False),
budou.Chunk(u'c', None, budou.TARGET_LABEL[2], True),
]
result = self.parser._concatenate_by_label(chunks, True)
self.assertEqual(
result, expected_forward_concat,
'Forward directional chunks should be concatenated to following '
'chunks.')
expected_backward_concat = [
budou.Chunk(u'ab', None, budou.TARGET_LABEL[0], True),
budou.Chunk(u'c', None, budou.TARGET_LABEL[2], True),
]
result = self.parser._concatenate_by_label(chunks, False)
self.assertEqual(
result, expected_backward_concat,
'Backward directional chunks should be concatenated to preceding '
'chunks.')
def test_group_chunks_by_entities(self):
# chunks: foo bar baz
# entity: ___ bar ___
chunks = budou.ChunkList(
[budou.Chunk('foo'),
budou.Chunk('bar'),
budou.Chunk('baz')])
entities = [{'beginOffset': 3, 'content': 'bar'}]
expected = ['foo', 'bar', 'baz']
result = self.parser._group_chunks_by_entities(chunks, entities)
self.assertEqual(expected, [chunk.word for chunk in result])
# chunks: foo bar baz
# entity: foo ba_ ___
chunks = budou.ChunkList(
[budou.Chunk('foo'),
budou.Chunk('bar'),
budou.Chunk('baz')])
entities = [{'beginOffset': 0, 'content': 'fooba'}]
expected = ['foobar', 'baz']
result = self.parser._group_chunks_by_entities(chunks, entities)
self.assertEqual(expected, [chunk.word for chunk in result])
def setUp(self):
chunks = budou.ChunkList()
chunks.append(budou.Chunk('ab', dependency=None))
chunks.append(budou.Chunk('cde', dependency=True))
chunks.append(budou.Chunk('fgh', dependency=False))
self.chunks = chunks
def test_get_source_chunks(self):
queue = self.parser._get_source_chunks(DEFAULT_SENTENCE_JA)
expected = [
budou.Chunk(u'六本木', label='NN', pos='NOUN', dependency=None),
budou.Chunk(u'ヒルズ', label='ADVPHMOD', pos='NOUN', dependency=None),
budou.Chunk(u'で', label='PRT', pos='PRT', dependency=False),
budou.Chunk(u'、', label='P', pos='PUNCT', dependency=False),
budou.Chunk(u'「', label='P', pos='PUNCT', dependency=True),
budou.Chunk(u'ご飯', label='DOBJ', pos='NOUN', dependency=None),
budou.Chunk(u'」', label='P', pos='PUNCT', dependency=False),
budou.Chunk(u'を', label='PRT', pos='PRT', dependency=False),
budou.Chunk(u'食べ', label='ROOT', pos='VERB', dependency=None),
budou.Chunk(u'ます', label='AUX', pos='VERB', dependency=False),
budou.Chunk(u'。', label='P', pos='PUNCT', dependency=False)
]
self.assertEqual(
[chunk.word
for chunk in expected], [chunk.word for chunk in queue.chunks],
'Words should be match between input text and retrieved chunks.')
self.assertEqual([chunk.dependency for chunk in expected], [
chunk.dependency for chunk in queue.chunks
], 'Dependency should be match between input text and retrieved chunks.'
)
def test_get_source_chunks(self):
tokens = [{
'dependencyEdge': {
'headTokenIndex': 1,
'label': 'NN'
},
'partOfSpeech': {
'tag': 'NOUN'
},
'text': {
'beginOffset': 0,
'content': u'六本木'
}
}, {
'dependencyEdge': {
'headTokenIndex': 8,
'label': 'ADVPHMOD'
},
'partOfSpeech': {
'tag': 'NOUN'
},
'text': {
'beginOffset': 3,
'content': u'ヒルズ'
}
}, {
'dependencyEdge': {
'headTokenIndex': 1,
'label': 'PRT'
},
'partOfSpeech': {
'tag': 'PRT'
},
'text': {
'beginOffset': 6,
'content': u'で'
}
}, {
'dependencyEdge': {
'headTokenIndex': 8,
'label': 'P'
},
'partOfSpeech': {
'tag': 'PUNCT'
},
'text': {
'beginOffset': 7,
'content': u'、'
}
}, {
'dependencyEdge': {
'headTokenIndex': 5,
'label': 'P'
},
'partOfSpeech': {
'tag': 'PUNCT'
},
'text': {
'beginOffset': 8,
'content': u'「'
}
}, {
'dependencyEdge': {
'headTokenIndex': 8,
'label': 'DOBJ'
},
'partOfSpeech': {
'tag': 'NOUN'
},
'text': {
'beginOffset': 9,
'content': u'ご飯'
}
}, {
'dependencyEdge': {
'headTokenIndex': 5,
'label': 'P'
},
'partOfSpeech': {
'tag': 'PUNCT'
},
'text': {
'beginOffset': 11,
'content': u'」'
}
}, {
'dependencyEdge': {
'headTokenIndex': 5,
'label': 'PRT'
},
'partOfSpeech': {
'tag': 'PRT'
},
'text': {
'beginOffset': 12,
'content': u'を'
}
}, {
'dependencyEdge': {
'headTokenIndex': 8,
'label': 'ROOT'
},
'partOfSpeech': {
'tag': 'VERB'
},
'text': {
'beginOffset': 13,
'content': u'食べ'
}
}, {
'dependencyEdge': {
'headTokenIndex': 8,
'label': 'AUX'
},
'partOfSpeech': {
'tag': 'VERB'
},
'text': {
'beginOffset': 15,
'content': u'ます'
}
}, {
'dependencyEdge': {
'headTokenIndex': 8,
'label': 'P'
},
'partOfSpeech': {
'tag': 'PUNCT'
},
'text': {
'beginOffset': 17,
'content': u'。'
}
}]
self.parser._get_annotations = MagicMock(return_value=(tokens, None))
chunks, _, _ = self.parser._get_source_chunks(u'六本木ヒルズで、「ご飯」を食べます。')
expected = [
budou.Chunk(u'六本木', label='NN', pos='NOUN', dependency=None),
budou.Chunk(u'ヒルズ', label='ADVPHMOD', pos='NOUN', dependency=None),
budou.Chunk(u'で', label='PRT', pos='PRT', dependency=False),
budou.Chunk(u'、', label='P', pos='PUNCT', dependency=False),
budou.Chunk(u'「', label='P', pos='PUNCT', dependency=True),
budou.Chunk(u'ご飯', label='DOBJ', pos='NOUN', dependency=None),
budou.Chunk(u'」', label='P', pos='PUNCT', dependency=False),
budou.Chunk(u'を', label='PRT', pos='PRT', dependency=False),
budou.Chunk(u'食べ', label='ROOT', pos='VERB', dependency=None),
budou.Chunk(u'ます', label='AUX', pos='VERB', dependency=False),
budou.Chunk(u'。', label='P', pos='PUNCT', dependency=False)
]
self.assertEqual(
[chunk.word
for chunk in expected], [chunk.word for chunk in chunks],
'Words should be match between input text and retrieved chunks.')
self.assertEqual([chunk.dependency for chunk in expected], [
chunk.dependency for chunk in chunks
], 'Dependency should be match between input text and retrieved chunks.'
)
def reset_queue(self):
chunks = [budou.Chunk('foo'), budou.Chunk('bar'), budou.Chunk('baz')]
queue = budou.ChunkQueue()
for chunk in chunks:
queue.add(chunk)
return queue
def test_get_source_chunks(self):
budou.api.get_annotations = MagicMock(
return_value=self.cases['ja-case1']['tokens'])
queue = self.parser._get_source_chunks(
self.cases['ja-case1']['sentence'])
expected = [
budou.Chunk(u'六本木', label='NN', pos='NOUN', dependency=None),
budou.Chunk(u'ヒルズ', label='ADVPHMOD', pos='NOUN', dependency=None),
budou.Chunk(u'で', label='PRT', pos='PRT', dependency=False),
budou.Chunk(u'、', label='P', pos='PUNCT', dependency=False),
budou.Chunk(u'「', label='P', pos='PUNCT', dependency=True),
budou.Chunk(u'ご飯', label='DOBJ', pos='NOUN', dependency=None),
budou.Chunk(u'」', label='P', pos='PUNCT', dependency=False),
budou.Chunk(u'を', label='PRT', pos='PRT', dependency=False),
budou.Chunk(u'食べ', label='ROOT', pos='VERB', dependency=None),
budou.Chunk(u'ます', label='AUX', pos='VERB', dependency=False),
budou.Chunk(u'。', label='P', pos='PUNCT', dependency=False)
]
self.assertEqual(
[chunk.word
for chunk in expected], [chunk.word for chunk in queue.chunks],
'Words should be match between input text and retrieved chunks.')
self.assertEqual([chunk.dependency for chunk in expected], [
chunk.dependency for chunk in queue.chunks
], 'Dependency should be match between input text and retrieved chunks.'
)
